io6 The Control of Mosquitoes 



As it was definitely proved that the forward long- 

 distance flight was of limited duration and occurred 

 just at dusk, it was assumed that there should be a 

 visible return flight at or before dawn. The first 

 morning this was carefully waited for from mid- 

 night to 6 A.M. but did not occur, and the observers 

 noticed very few mosquitoes, although thousands 

 had passed them on the forward flight. It was 

 thought that there must be a return flight to the 

 propagation area of a nature not yet understood. 

 It had been noted that at the start of the forward 

 flight {i.e. over the canal) the Anopheles were forty 

 feet up in the air, and it was possible that the end 

 of the return flight was also above the range of 

 vision. Additional investigation, however, showed 

 that there was a marked return flight and though 

 this happened on several successive mornings, other 

 mornings showed practically no return fHght from 

 the village, or houses, to the breeding place, so far 

 as could be noticed. This return flight did not 

 begin until 6.00 a.m. although there was sufficient 

 light to read by ten minutes before that time. 

 The return flight, once fairly started, was of much 

 shorter duration and more rapid than the forward 

 flight. The "night-jars" accompanied the return 

 flight, but were absent on those mornings when no 

 return flight occurred. As the daylight became 



